Sabitlenmiş Tweet
The universe is weird as shit and I'm profoundly grateful that it permits us to poke at it.
English
Brydon Eastman
1.1K posts
Brydon Eastman
@brhydon
🇨🇦 Mathematician (heavy on the ish) @thinkymachines Prev. MTS @OpenAI; PhD @WaterlooMath Certified wife guy, featured twice in Lego Magazine © ☕//🤔➡️💻
Katılım Mayıs 2018
1.2K Takip Edilen3.3K Takipçiler
one thing about atom egoyan is that he will always torontomaxx
Andrew Frye@itsandrewfrye
Chloe was really Torontomaxxing here
English
@theojaffee I mean, the dot quebec is definitely for the *province* of quebec, not specifically Quebec City.
English
Brydon Eastman retweetledi
Brydon Eastman retweetledi
Model has good taste and so do the modellers (taste in training infra, that is)
Tinker@tinkerapi
tasty AND trained on Tinker >>>
English
Brydon Eastman retweetledi
Probably the best part of grad school was the free pie, and I'm still raking in the benefits years later.
Thanks for the free pie @WaterlooMath ! I promise to *consider* sharing it
English
@hallerite claude code but for powerlifting, who's building this ??????
English
@rapha_gl "I'm so Important that a team of Swiss experts with knowledge passed down thru generations slaved away to make me this watch that's practically indistinguishable, or inferior to, the ones from the robofactory, but mine was made by PEOPLE ergo I'm better than you or something"
English
@rapha_gl idk I'm skeptical, I can totally see a future where you can get a AP from a robofactory crafted to be indistinguishable from the royal oaks of today and there still being a long tail of demand for luxury hand crafted watches if only for the emotional resonance and flex
English
English
Brydon Eastman retweetledi
@PrimeIntellect shares our mission of democratizing AI-training and has built a stack for RL including the Verifiers library. Verifiers’ integration with Tinker lets builders tap into Prime’s growing Environments Hub ecosystem, or create their own.
x.com/willccbb/statu…
will brown@willccbb
u can use the Environments Hub with the thinky product btw github.com/thinking-machi…
English
As a species we've been bad at disentangling intelligence, sentience, and agency.
If you think that Sam has divine souls behind a tap that he only opens if you pony up $20, ofc you'd think that's horrific, but i refuse to be pessimistic about problem solving too cheap to meter.
Autism Capital 🧩@AutismCapital
Congratulations. You’ve lived long enough to see man made horrors beyond comprehension.
English
@ChowdhuryNeil It's a really good in context reasoning engine! I also completely disagreed with the original tweet.
English
i disagree. gpt-oss-120b is *the* model i use most frequently for my research. it is ridiculously good for how cheap it is (5b active). it gets hate for being worse than larger chinese models, but it is one of my favorites -- i really hope that openai releases future oss models
BURKOV@burkov
GPT-OSS-120B is a useless model. Nothing I tried to use it for worked as you would expect from a model of this size. It doesn't respect the constraints, performs poorly when the task description precedes the text the task is supposed to apply to, it stops the generation at random moments without finishing the sentence, and it generates repetitive expressions that don't ever stop. All these are properties of 7B-parameter models of late 2023. IMO, altman released this model to avoid being accused of lying after he made a drunk promise to release a competitive open-weight model.
English
Brydon Eastman retweetledi